Key Words: Alkaline phosphatase isoenzymes; Alanine aminotransferase; Abdominal obesity; ... bone-kidney ALP), intestinal, placental, and germ cell.9-11 As.
Clinical Chemistry / ANTHROPOMETRY AND ALKALINE PHOSPHATASE
The Relationship Between Anthropometry and Serum Concentrations of Alkaline Phosphatase Isoenzymes, Liver Enzymes, Albumin, and Bilirubin Aus T. Ali, PhD, Janice E. Paiker, MBBCh, and Nigel J. Crowther, PhD Key Words: Alkaline phosphatase isoenzymes; Alanine aminotransferase; Abdominal obesity; Liver function; Levamisole DOI: 10.1309/9N346GXX67B6PX5W
Abstract Alkaline phosphatase (ALP) is present in human preadipocytes. The aim of this study was to determine the relationship between anthropometry and serum levels of ALP isoenzymes, liver enzymes, albumin, and bilirubin. Anthropometric variables; serum total, bone, liver, and intestinal ALP levels; and alanine aminotransferase (ALT), albumin, total protein, total bilirubin, and γ-glutamyltransferase serum levels were measured in 100 volunteers. The levels (given as median [interquartile range]) for total (74.0 U/L [30.0 U/L] vs 62.0 U/L [22.0 U/L]; P < .05) and liver ALP (37.3 U/L [14.6 U/L] vs 26.1 U/L [12.0 U/L]; P < .05) were higher in obese than in lean subjects. The levels of the other ALP isoenzymes and blood analytes were not significantly different between these groups. Albumin and ALT were the only blood proteins studied with serum levels that correlated significantly with waist circumference. This present study demonstrates a relationship between abdominal obesity and serum ALT levels and between body mass index and ALP levels. These findings suggest that serum ALP, particularly liver ALP, is derived from adipose and hepatic tissue.
Besides its well-known association with chronic diseases of lifestyle,1,2 obesity has been identified as an important contributing factor to raised serum levels of hepatic enzymes.3-5 It is thought that this relationship is due to the high release of free fatty acids from the visceral fat depot into the portal circulation, leading to nonalcoholic steatohepatitis (NASH).6,7 NASH is considered a characteristic of fat distribution that is associated strongly with the metabolic syndrome in obese8 and nonobese subjects.6 Alkaline phosphatase (ALP) is a membrane-bound enzyme found in a wide variety of tissues, including liver. There are 4 ALP isoenzymes in humans, each coded by a separate gene: tissue nonspecific (TNALP; also known as liverbone-kidney ALP), intestinal, placental, and germ cell.9-11 As its name implies, ALP works in an alkaline environment, suggesting that the enzyme, although present in blood, is inactive in this environment. The enzyme is known to have phosphoprotein phosphatase and transphosphorylation activity and might have an important role in bone mineralization.12 However, its function in other tissues is not known. The serum levels of liver and bone ALP are used widely in the diagnosis of hepatobiliary disease and various bone disorders,13,14 respectively. It recently was reported that the TNALP isoenzyme is present in human and murine preadipocytes and might have a role in the intracellular accumulation of triglycerides that characterizes the process of adipogenesis.15,16 The existence of ALP in human preadipocytes is of interest because it is conceivable that adipose tissue might be a source of serum ALP. Furthermore, the positive relationship between measures of abdominal obesity and serum liver enzyme levels demonstrates that adipose tissue mass also can influence the release of liver products into the circulation. Thus, Am J Clin Pathol 2006;126:437-442
© American Society for Clinical Pathology 437
DOI: 10.1309/9N346GXX67B6PX5W
437 437
Ali et al / ANTHROPOMETRY AND ALKALINE PHOSPHATASE
the level of TNALP in serum might be influenced by total and abdominal adipose tissue mass. The purpose of the present study was to analyze the relationships between ALP isoenzyme, liver enzyme, albumin, and bilirubin serum concentrations and measures of total and abdominal adipose tissue mass.
Materials and Methods Food Intake and Total Serum ALP Concentration The effect of food intake on serum total ALP levels was studied to determine whether nonfasting blood samples could be used for the measurement of ALP. A total of 12 volunteers (10 women) were used, and ALP levels were measured in venous serum samples after an overnight fast and 60 minutes after breakfast. The Use of Levamisole in Estimating Serum Intestinal ALP Concentration Levamisole is a specific inhibitor of the TNALP isoenzyme.17 Therefore, this inhibitor was used to block activity of this enzyme in serum samples, the remaining activity being attributable to intestinal ALP. The level of levamisole required for maximal inhibition of TNALP activity was assessed by adding levamisole to serum samples obtained from 4 healthy male volunteers to give final concentrations of 0, 8, 16, 32, and 64 mmol/L. Study Subjects, Anthropometric Measurements, and Blood Sampling The study subjects comprised 100 black Africans (26 men) of varying body mass index (BMI) values. Ethical approval for the collection of serum samples was obtained from the University of Witwatersrand (Johannesburg, South Africa), Faculty of Health Sciences Human Ethics Committee. Informed consent was obtained from each volunteer. Only subjects who had no known liver pathology or bone fractures within the last 2 years were included in the study. Weight and height were measured by using a stadiometer (Modern Scale, Johannesburg). The waist/hip ratio (WHR) was measured by taking waist circumference as the midpoint between the lower rib margin and the iliac crest and hip circumference as the widest circumference of the buttock. The subjects were split into 3 groups according to BMI, ie, lean (BMI, 24.9 and 29.9). Women also were divided into 2 groups based on age, using a cutoff age of 50 years to denote premenopausal and postmenopausal status.18 A 5-mL, nonfasting blood sample was obtained from the antecubital vein of the forearm from each subject. The sample 438 438
Am J Clin Pathol 2006;126:437-442 DOI: 10.1309/9N346GXX67B6PX5W
was centrifuged immediately and the serum removed and stored at –20°C until assayed. Estimation of Serum ALP Isoenzyme Concentrations Bone ALP was measured by using a bone-specific ALP enzyme immunoassay (Quidel, San Diego, CA). Serum levels of intestinal ALP were estimated by measuring ALP activity in serum treated with 32 mmol/L of levamisole. Liver ALP levels were calculated by subtracting the sum of the bone and intestinal ALP concentrations from the total serum ALP activity measured using a colorimetric assay (ALP IFCC liquid assay, Roche Diagnostics, Mannheim, Germany) performed on a Modular autoanalyzer (Roche Diagnostics). Measurement of Serum Concentrations of Liver-Derived Proteins Serum levels of albumin, total bilirubin, alanine aminotransferase (ALT), total protein, and γ-glutamyltransferase were measured using commercial kits (Roche Diagnostics) performed on the Modular autoanalyzer using standard procedures. The reference ranges for these analytes were obtained from the kit manufacturers. Statistical Analyses All study variables displayed skewed distributions and, therefore, were transformed to normality by taking log, reciprocal, or squared values. Data in tables and the text are expressed as median (interquartile range) unless otherwise stated. Comparisons between sexes were performed by analysis of covariance (ANCOVA) with adjustment for age, BMI, and WHR or waist circumference. Subjects were split into 3 groups according to BMI and data compared across groups using ANCOVA adjusted for age, sex, and WHR or waist circumference. The analysis of specific differences between 2 BMI groups was performed by using the Tukey post hoc test for unequal numbers. The relationship between age and serum ALP isoenzyme activity was analyzed separately in men and women by using Pearson regression. Differences in means between premenopausal and postmenopausal women were analyzed by using analysis of variance. Multiple regression analysis was used to determine the relationship between serum peptide levels and measures of abdominal obesity, ie, WHR and waist circumference.
Results The Effect of Food Intake on Serum Total ALP Concentration The serum total ALP level (mean ± SD) was 61.4 ± 17.6 U/L before breakfast and 60.8 ± 17.1 U/L after breakfast © American Society for Clinical Pathology
Clinical Chemistry / ORIGINAL ARTICLE
(P = .43). Thus, food intake had no effect on serum ALP levels, and, therefore, in the larger study, nonfasting blood samples were used to determine serum levels of ALP. The Use of Levamisole for Estimating Serum Intestinal ALP Concentrations Maximal inhibition of ALP activity occurred at 32 mmol/L of levamisole, producing an inhibition of 85%. Therefore, the estimation of intestinal ALP activity was performed using serum samples treated with 32 mmol/L of levamisole. Comparison of Male and Female Study Subjects The study comprised 26 male and 74 female black African volunteers. The women were older (39 [15] vs 32.5 [6] years; P = .002) and had a higher BMI (30.6 [11.1] vs 21.9 [4.4]; P < .0001, corrected for age) and a lower WHR (0.81 [0.13] vs 0.88 [0.05]; P = .002, corrected for age and BMI) than the men. The men had higher serum albumin levels (46 g/L [3 g/L] vs 42 g/L [3 g/L]; P < .0001, corrected for age, BMI, and WHR) and lower total protein levels (76 g/L [4 g/L] vs 79 g/L [10 g/L]; P = .004 corrected for age, BMI, and WHR) than the women. (Albumin and protein values are given in Système International units; to convert to conventional units [g/dL], divide by 10.0.) Similar trends were observed when ANCOVAs were performed with waist circumference replacing WHR as an independent variable. The ALT levels were significantly higher in men (15.5 U/L [5 U/L]) than in women (12.5 U/L [6 U/L]; P = .04, corrected for age and BMI), but this difference became nonsignificant
(P = .35) when WHR or waist circumference was included as an independent variable in the ANCOVA. The Effect of BMI on Serum ALP Activity and Liver Enzyme, Albumin, and Bilirubin Concentrations The data in ❚Table 1❚ demonstrate that obese subjects were significantly older than lean subjects and had a higher waist circumference than all the other subject groups. Furthermore, the obese group had significantly higher total and liver ALP levels than the lean group after adjustments for age, sex, and WHR or waist circumference. Multivariate regression analysis using sex, age, and WHR as independent variables ❚Table 2❚ demonstrated that ALT levels were determined predominantly by WHR. The main determinant of albumin levels was found to be sex, with a very weak input from WHR. Total and liver ALP concentrations correlated positively with age but not with WHR. This also was the case for bone (β coefficient = .34; P = .001 with age) and intestinal (β coefficient = .26; P = .015 with age) ALP. None of the other ALP isoenzyme or liver enzyme, albumin, or bilirubin serum levels showed significant correlations with WHR. When waist circumference was used as an independent variable in place of WHR, relationships similar to those shown in Table 2 were observed. Effect of Age on Serum ALP Isoenzyme Activity Serum levels of total (r = 0.44; P < .0001), bone (r = 0.42; P < .0001), liver (r =0.31; P = .008), and intestinal (r = 0.28; P = .014) ALP correlated positively with age in women, whereas only liver ALP (r = 0.48; P = .013) correlated with age in men.
❚Table 1❚ Anthropometric Measurements and ALP, Liver Enzyme, and Liver Biosynthetic Product Concentrations in Lean, Overweight, and Obese Subjects* Variable Age (y) BMI Waist (cm) WHR Total ALP (U/L) Liver ALP (U/L) Bone ALP (U/L) Intestinal ALP (U/L) ALT (U/L) GGT (U/L) Total protein (g/L) Albumin (g/L) Total bilirubin (µmol/L)
Lean (n = 37) 33.0 (10.0) 21.9 (3.6) 74.0 (8.0) 0.82 (0.10) 62.0 (22.0) 26.1 (12.0) 28.2 (12.0) 4.0 (3.0) 14.0 (6.0) 22.0 (17.0) 76.0 (9.0) 43.0 (5.0) 7.0 (4.0)
Overweight (n = 22)
Obese (n = 41) 42.0 (16.0)† 35.5 (8.6)‡§ 98.5 (14.5)‡§ 0.82 (0.16) 74.0 (30.0)|| 37.3 (14.6)|| 32.6 (21.9) 5.0 (4.0) 12.0 (6.0) 25.0 (34.0) 76.0 (6.0) 42.0 (3.0) 6.0 (4.0)
37.5 (14.0) 27.8 (10.6)‡ 86.7 (10.5)‡ 0.85 (0.08) 66.5 (21.0) 29.3 (15.2) 30.9 (11.1) 4.0 (5.0) 14.5 (6.0) 24.5 (48.0) 79.0 (10.0) 43.0 (4.0) 6.5 (4.0)
ALP, alkaline phosphatase; ALT, alanine aminotransferase; BMI, body mass index; GGT, γ-glutamyltransferase; WHR, waist/hip ratio. * Values are expressed as median (interquartile range) in Système International units. Conversions to conventional units are as follows: total ALP, ALT, and GGT (U/L), divide by 1.0; total protein and albumin (g/dL), divide by 10.0; total bilirubin (mg/dL), divide by 17.1. † P < .005 vs lean group. ‡ P < .0005 vs lean group. § P < .0005 vs overweight group. || P < .05 vs lean group.
Am J Clin Pathol 2006;126:437-442
© American Society for Clinical Pathology 439
DOI: 10.1309/9N346GXX67B6PX5W
439 439
Ali et al / ANTHROPOMETRY AND ALKALINE PHOSPHATASE
❚Table 2❚ Multiple Regression Analyses* Model No.
Dependent Variable
1
ALT (log)
2
Albumin (square root)
3
Total ALP (log)
4
Liver ALP (log)
Independent Variables
β-Coefficient (P)
Adjusted R2 (P)
Sex Age (log) WHR (reciprocal) Sex Age (log) WHR (reciprocal) Sex Age (log) WHR (reciprocal) Sex Age (log) WHR (reciprocal)
.15 (.15) –.006 (.95) –.37 (.0002) .48 (
